BsR: Evaluating the Metric Behind Baseball’s Runners



A player’s ability to run the bases with given innate baseball knowledge and speed has always been considered when evaluating their overall skill. Although more heavily weighted in the past versus the present, this asset in a player’s repertoire can still add several runs to a team’s overall totals. As baseball evolves, the main question turns into how this skill can be measured by an objective, quantitative number. Introducing Base Running (or BsR for short). The main objective of BsR is to combine the many facets of running the bases into a single statistic, allowing a fan or team to compare two players. Knowing the scale of the one number measuring Base Running is extremely important, but knowing all of the factors involved allows fans to make a well-informed decision when evaluating a player and the worthiness of the actual stat. This includes looking at the equation for BsR and examining the value that it offers.


While the statistic itself is an uncomplicated sum of running values, the intricacies behind those values are fairly complicated. To properly break this down, let’s start with the equation itself.


BsR = wSB + UBR + wGDP


As I stated, not very complicated. The first value, wSB, stands for Weighted Stolen Base Runs. This calculates the run value of a player’s ability to steal bases. The equation for this formula is quite long and includes several factors that are necessary to know. They are as follows: runSB (Run Value of Stolen Bases), runCS (Run Value of Caught Stealing), and lgwSB (League Stolen Base Runs). These are all measured through the run expectancy matrix (RE24), which measures the given run probability in each of the 24 base-out states in the game of baseball.


The second value, UBR, stands for Ultimate Base Running. Produced by the same company that invented UZR (Ultimate Zone Rating) for defense, UBR averages and weighs the expected run values that a player produces by advancing bases and compares it to the average of his peers when similar situations arise.


The last value, wGDP, stands for Weighted Grounded into Double Play Runs. As the name properly states, it is the weighted value of runs a player produces by avoiding or hitting into double plays.


Each baseball equation needs to accurately measure the difference a player makes. BsR, for the most part, clearly does this. The only ambiguity lies in UBR. The calculation of UBR requires a unique primer over a simple equation (the equation shown below is the rough function of the primer). And while generally this cloak is not preferred, base running is highly difficult to calculate - similar to defensive characteristics. Although, when examining the Ultimate Base Running (UBR) primer, the entire statistic is primarily focused on the Run Expectancy of a given situation. And while it does serve its purpose, it doesn’t allow for predictive measurements to be made. Because it lacks Statcast considerations of factors like Sprint Speed, Jump-Time, etc., a player’s given Base Running score will not correlate very well to his future numbers. As the underlying factors stand to be the reasoning behind the production, not considering such metrics leads to a more baseless sense of judgment.


All things considered, I deem the equation as somewhat satisfactory in measuring produced value - leaving room for expected value. To give a more clear picture of the aforementioned references, here are the equations for the values.


wSB = SB * runSB + CS * runCS – lgwSB * (1B + BB + HBP – IBB)


UBR = (Runner’s RV - xRV in Given Play) * Runner’s Plays


wGDP = GDP RV - GDP xRV / DP Opportunities


As with any statistic, to truly be informed, a fan needs to know of its value proposition. BsR is denominated into runs, with any number above or below 0 revealing the number of runs a player has added or subtracted from his team based on his baserunning alone. According to Fangraphs, on a single-season basis, a score of 0 is considered average while 8 is considered excellent. A similar logic applies in the negatives, with -6 to -8 being considered atrocious. And while the 0 may seem to be a constant, it is an ever-changing calculation from season to season. Player X may run the bases even better than last year in terms of nominal value, but including the increased average from last year, his BsR may look worse. While seemingly a flaw, nominal value often fails to tell the whole story, which is what adjusting for averages attempts to do. The actuality of how well this is done is up to the avid fan, as I’d imagine teams already use a much more advanced model.


Speaking objectively, it is worth noting the amount of impact, or value, this factor has on the entirety of a team. Remembering that 10 runs roughly equates to 1 win, the scale fails to touch a win added over a 162-game season. Even if playing tremendously, it is very unlikely that a player could gain an extra win (only Starling Marte did this in 2021), or lose one in the adverse. In a game of inches, every run does matter. But for how much stock teams put into incredibly fast players that lack necessary tools (Ex. Billy Hamilton), it may be an over-consideration. With hitting being arguably the most efficient way to produce value, a slow slugger is a safer bet in overall value than a fast, mediocre hitter. That should not be a new lesson, as time after time, incredibly skilled base runners have left the league due to underperformance while the unskilled baserunning sluggers remain. With this in mind, there should be a reconsideration in weighing BsR. Understanding its gauge of actual value can affect the decisions made in pointing out a flaw or a pro in a given player.


Base Running (BsR), in considering the various factors present, is an ultimately valid statistic that provides a solid method to objectively value base running. But as mentioned earlier, there is room for an upgrade, depending on your view of the function of the stat. In my opinion, these types of tell-all numbers should weigh for Statcast considerations of the individual player based on the league average. If one believes BsR should only be the actual value produced, that is respectable - this measurement tells that story accurately. But if one, like me, believes that the future numbers should correlate better to past numbers in evaluating actual value, then expected values need to be weighted. In truth, a hybrid version could be created, or an xBsR alone. This is only being suggested as I believe the statistic fails to tell the whole story, similar to other regular production numbers. Either way, BsR’s actual value is little in comparison to the other skillsets. As I emphasized in the paragraph above, too many teams have fallen into the fatal mistake of overvaluing the seemingly supernatural skill. It is still worth considering, but not on a make-or-break basis. I would argue that almost all advanced statistics based on aspects of the game should be considered equally - not weighted equally. And while a lovely solution to a needed problem, BsR should rank on the lower scale of importance when evaluating a player.


Sources:


Baseball-Reference.com

Baseballsavant.com

Fangraphs.com